﻿Hydrogen through a Palladium Septum. 211 



of the pressure then changed, and the usual rise with hydrogen 

 and fall with air occurred. I think that this is to be attri- 

 buted to the fact that the palladium vessel increases its capacity 

 suddenly on exposure to hydrogen, and decreases it on ex- 

 posure to air after having been exposed to hydrogen. But 

 the total change of volume is a very minute one ; it is notice- 

 able only by the rise or fall of the mercury in the capillary 

 tube to the amount of one or two millimetres ; and the total 

 change of volume is certainly less than 0*1 per cent, of the 

 capacity of the vessel. 



Experiments 3, 4, and 5. — The next three experiments gave 

 readings much lower than those previously obtained. The 

 barometric reading was 748*6 millim. The rise in experi- 

 ment 3 was 650*3 millim. ; in experiment 4, 654*2 millim. ; 

 and in experiment 5, 657*8 millim. 



This led me to suspect that the palladium tube had under- 

 gone alteration. It was therefore disconnected and cleaned, 

 as before described. It was finally filled with nitrogen. 



Experiments 6, 7 , 8, 9, and 10. — The barometric pressure 

 was 770*2 millim. The final pressure in experiment 6 was 

 698*5; in 7, 692*1; in 8, the barometric pressure being 768*8 

 millim., the final pressure was 686*4 millim. In experiment 

 9, with a barometer-reading of 762*2 millim., the reading- 

 was 692*4 millim. ; and in 10, the barometer-reading being 

 759*2, the pressure was 688*4 millim. 



To summarize these results it is necessary to compare the 

 pressure of the internal hydrogen with that of the external 

 hydrogen. The following table gives their ratios : — 



Experiment. (1) (2) (3) (4) (5) 



Ratio .... 0*9400 0*9798 0*8687 0*8739 0*8780 



Experiment. (6) (7) (8) (9) (10) 



Ratio .... 0*9069 0*8986 0*8928 0*9084 0*9067 



These results do not exhibit any great degree of constancy, 

 but I am sure that their difference depends largely, if not 

 wholly, on the state of the surface of the palladium. The 

 mean result is 0*9053, a number closely approximating to 

 the last five determinations. This would lead to the con- 

 clusion that the pressure of the internal hydrogen rises to 

 rather more than A of the pressure of the external hydrogen 

 at 280°. J ° 



Experiment 11. — The temperature was now kept at the 

 boiling-point of quinoline (about 237°). The passage of hy- 

 drogen through palladium is so slow at this temperature that 



P2 



